Mouse mismatch repair gene Msh2 is not essential for ... - Nature

ã

Oncogene (2001) 20, 538 ± 541 2001 Nature Publishing Group All rights reserved 0950 ± 9232/01 $15.00 www.nature.com/onc

Mouse mismatch repair gene Msh2 is not essential for transcription-coupled repair of UV-induced cyclobutane pyrimidine dimers Edwin Sonneveld1,2, Harry Vrieling2,3, Leon HF Mullenders*,2,3 and Anneke van Ho€en2 1

Department of Cell Biology and Genetics-MGC, Erasmus University Rotterdam, P.O. Box 1738, 3000 DR Rotterdam, The Netherlands; 2Department of Radiation Genetics and Chemical Mutagenesis-MGC, Leiden University Medical Center, Leiden, The Netherlands; 3J.A. Cohen Institute, Inter-University Institute for Radiopathology and Radiation Protection, Leiden, The Netherlands

The human mutS homolog gene MSH2 is essential for DNA mismatch repair (MMR) and defects in this gene can result in increased mutagenesis, genomic instability and hereditary nonpolyposis colorectal cancer (HNPCC). Besides correcting mismatch errors arising from DNA replication, it was shown that de®ciencies in bacterial and human MMR genes including MSH2 resulted in defective transcription-coupled repair (TCR) of UV-induced photolesions. Here we show that MMRde®cient ®broblasts derived from two independent isogenic mouse strains with de®ned Msh2 de®ciencies are as pro®cient in TCR of UV-induced cyclobutane pyrimidine dimers (CPD) as wildtype ®broblasts. Our results indicate that in mouse cells Msh2 is not essential for TCR of UV-induced CPD in contrast to bacteria and human cells and suggest that the biological e€ects of UV in mouse Msh27/7 cells and mice are not due to defective TCR. Oncogene (2001) 20, 538 ± 541. Keywords: mismatch repair; transcription-coupled repair; UV-damage; mouse ®broblasts

To counteract the deleterious e€ects of various types of DNA damage caused by exogenous as well as endogenous agents, mammalian cells are equipped with multiple DNA repair pathways. DNA lesions as well as mismatches in DNA are, if not properly repaired, considered to be responsible for a majority of human cancers (Friedberg et al., 1995). The general pathway involved in the repair of mismatches resulting from replication errors in DNA is the mismatch repair (MMR) pathway (reviewed by Buermeyer et al., 1999). Mutations in several human MMR genes (e.g. MSH2, MSH6, MLH1 and PMS2) have been associated with hereditary predisposition to cancer (hereditary nonpolyposis colorectal cancer (HNPCC)), and a subset of sporadic cancers (Liu et al., 1995; Lynch et al., 1997).

*Correspondence: LHF Mullenders, Department of Radiation Genetics and Chemical Mutagenesis-MGC, Leiden University Medical Center, P.O. Box 9503, 2300 RA Leiden, The Netherlands Received 18 October 2000; revised 15 November 2000; accepted 16 November 2000

The possibility exists that this predisposition to cancer is not only caused by a failure of MMR de®cient cells to repair mismatches during DNA replication, but also by a failure of damaged and mutation-prone cells to trigger MMR-mediated apoptosis (Wu et al., 1999). Another major DNA repair pathway is nucleotide excision repair (NER) (reviewed by de Laat et al., 1999) dedicated to the removal of DNA helix distorting lesions including ultraviolet (UV) light-induced cyclobutane pyrimidine dimers (CPD). Defects in human NER are manifested in the rare hereditary human disorder xeroderma pigmentosum (XP) associated with UV-sensitivity, predisposition to skin cancer and neurological abnormalities (Bootsma et al., 1998). NER exists of two subpathways: transcription-coupled repair (TCR) responsible for the preferential removal of DNA damage from the transcribed strand (TS) of active genes, and global genome repair (GGR) involved in repair of lesions over the entire genome including the nontranscribed strand (NTS) of active genes. Factors speci®c for proper activity of TCR in mammalian cells have been identi®ed including the Cockayne syndrome gene products CSA and CSB (van Ho€en et al., 1993), Brca1 (Gowen et al., 1998) and MMR genes (Mellon et al., 1996; Leadon and Avrutskaya, 1997). Defects in MMR genes have been reported to result in defective TCR of UV photoproducts and oxidative damage. In human colon cancer cells and bacteria, defects in MMR (MLH1, MSH2 and their bacterial homologs mutL and mutS) have been shown to abolish TCR of UV-induced photolesions (Mellon et al., 1996; Mellon and Champe, 1996; Leadon and Avrutskaya, 1997), whereas MMR defective yeast cells were una€ected in TCR of UV photolesions (Sweder et al., 1996; Leadon and Avrutskaya, 1998). In addition, MMR de®ciency appeared to impair TCR of oxidative damage in human tumor cells and yeast (Leadon and Avrutskaya, 1997, 1998). Furthermore, the MMR protein Msh2 has been shown to exist in a complex with various NER proteins, suggesting that these proteins function in common processes (Bertrand et al., 1998). Until now the involvement of MMR in TCR in mammalian cells was shown only in human tumor cell lines (Mellon et al., 1996; Leadon and Avrutskaya, 1997). In recent studies we (M van Oosten, unpublished results) and others (E Friedberg, personal

Transcription-coupled repair in mismatch repair deficient mouse cells E Sonneveld et al

communication) noticed that Msh2 de®cient mice exhibit UV sensitivity (edema, erythema) indistinguishable from that in isogenic wildtype mice. This phenotype of Msh27/7 mice is inconsistent with a defect in TCR of UV-induced photolesions, since mice with defective TCR appear to be 10-fold more sensitive to the induction of erythema/edema and apoptosis (Berg et al., 2000; van Oosten et al., 2000). Regarding the UV sensitivity of MMR de®cient human cell lines (including MSH2 de®cient LoVo cells), con¯icting observations have been reported: Mellon et al. (1996) showed that MMR de®cient cells were more sensitive to the killing e€ects of UV irradiation than MMR pro®cient cells, whereas Leadon and Avrutskaya (1997) demonstrated no increased sensitivity of MMR de®cient human cell lines towards UV. Furthermore, mouse embryonic ®broblasts (MEFs) derived from Msh27/7 mice showed normal sensitivity to UV (Reitmair et al., 1997), so the question of whether Msh2 has a role in mouse NER remains unanswered. These observations prompted us to investigate the role of Msh2 in TCR of CPD in mouse cells derived from wildtype and two isogenic mouse strains with de®ned Msh2 de®ciency. Firstly, wildtype as well as Msh2+/7 and Msh27/7 primary mouse dermal ®broblasts (MDFs) derived from newborn mice with a deletion in Msh2 exon 7 (Smits et al., 2000) were isolated. Furthermore, adenovirus E1A immortalized wildtype and Msh27/7 MEFs (Berry et al., 2000) derived from embryos with a hygromycin insertion between codons 588 and 589 of Msh2 (de Wind et al., 1995) were used. To determine Msh2 protein levels in the various cell lines, cell lysates from MEFs and MDFs were subjected to Western blot analysis (Figure 1), employing a rabbit polyclonal antibody directed against full-length Msh2 (de Wind et al., 1998). Embryonic stem (ES) cells served as positive control, since these cells express high levels of Msh2 (de Wind

Figure 1 Expression of Msh2 protein in wildtype and Msh27/7 mouse cells. Cell lysates from E14 (129/Ola) derived ES cell line IB10.51 (de Wind et al., 1999), adenovirus E1A immortalized wildtype and Msh27/7 MEFs (Berry et al., 2000) derived from wildtype and Msh27/7 mouse embryos (de Wind et al., 1995), and primary wildtype, Msh2+/7 and Msh27/7 MDFs (derived from Msh2D7N newborn mice; Smits et al., 2000) were prepared in Laemmli bu€er. The lysates were subjected to Western blotting analysis using a rabbit polyclonal antibody raised against fulllength mouse Msh2 protein (de Wind et al., 1998). Msh2 protein is indicated with an arrowhead. The upper band represents an aspeci®c signal of the used antibody and serves as control for equal protein loading

et al., 1999). Msh2 protein could be detected in ES cells, wildtype MEFs, wildtype MDFs and MDFs heterozygous for Msh2, but the Msh2 protein was absent in the MEFs and MDFs homozygous for the Msh2 mutation, indicating functional inactivation of the Msh2 gene in cells derived from both types of knock-out mice (Figure 1).

539

Figure 2 Repair of CPD in primary mouse dermal ®broblasts. Representative autoradiograms (a) and graphs (b and c) showing the removal of CPD from the transcribed strand (TS) and nontranscribed strand (NTS) of the active p53 and Hprt genes in primary MDFs derived from wildtype and Msh27/7 newborn mice (Smits et al., 2000) irradiated with 10 J/m2 UV. CPD frequencies in the 16 kb EcoRI restriction fragment of the p53 gene and the 18 kb EcoRI restriction fragment containing exon 6 ± 9 of the Hprt gene were measured as described previously (Van Ho€en et al., 1995; Van Sloun et al., 1999). Brie¯y, equal amounts of DNA were either treated or mock treated with the CPD-speci®c enzyme T4 endonuclease V. The samples were subjected to electrophoresis in an alkaline 0.6% agarose gel and transferred to a nylon membrane (Hybond N+, Amersham) which was hybridized with a radiolabeled strand-speci®c probe. Autoradiograms were made and full-size fragments were quanti®ed by scanning of the ®lters using the InstantImagerTM (Packard Instrument Company, Meriden, CT, USA). (b) CPD frequencies were calculated by comparing the density of the band in the T4 endonuclease treated sample with that in the mock treated sample, assuming a Poisson distribution of lesions. Msh27/7 MDFs (*, *); wildtype MDFs (&, &); TS (closed symbols), NTS (open symbols) of p53 (b) and Hprt (c). Data points are means of at least two independent experiments and are presented as mean+s.e.m. Oncogene

Transcription-coupled repair in mismatch repair deficient mouse cells E Sonneveld et al

540

Oncogene

TCR was ®rstly measured in UV-exposed con¯uent growing primary MDF cells derived from the skin of newborn mice. We assessed the removal of CPD from the TS and NTS of p53 and Hprt in wildtype and Msh27/7 MDFs after exposure to 10 J/m2 UV light (Figure 2a,b). Both wildtype and Msh27/7 MDFs showed fast repair of CPD in the TS of p53: 6 h after UV irradiation already 80% of the UV-induced CPD were removed while repair was complete after 24 h, as previously shown for normal MDFs (Cheo et al., 1997). However, CPD were hardly removed from the NTS of p53 (Figure 2a,b). The kinetics for the removal of CPD in the Hprt gene appeared to di€er from those observed for the p53 gene. In wildtype MDFs only 50% of the UV-induced CPD were removed from the TS of Hprt after 24 h, whereas in Msh27/7 MDFs almost all CPD were repaired after 6 h (Figure 2a,c). This di€erence in repair kinetics of the Hprt gene between wildtype and Msh27/7 MDFs most likely relates to the fact that wildtype MDFs were isolated from a female mouse containing two copies of the Xchromosomal Hprt gene per cell, while the Msh27/7 MDFs were derived from a male mouse containing only one copy per cell. From the two Hprt gene copies in female cells only one is actively transcribed resulting in 50% reduction of TCR which can be noticed for the wildtype MDFs (Figure 2c). Taken together, the data indicate that TCR of CPD in the Hprt and p53 genes is similar for Msh2 de®cient MDFs and Msh2 pro®cient MDFs. This result contrasts the de®cient TCR or UVinduced CPD in human LoVo tumor cells lacking MSH2 (Mellon et al., 1996). Since human keratinocytes immortalized with human papillomavirus show impaired TCR of UV-induced CPD compared with parental primary keratinocytes (Rey et al., 1999) the di€erence in TCR between human LoVo tumor cells and Msh2 de®cient MDFs might be caused by the use of primary cells. In order to investigate the possibility that the transformed state of cells could interfere with TCR of CPD, we also examined TCR in adenovirus E1A immortalized MEFs. In exponentially growing wildtype MEFs, CPD in the TS were repaired more rapidly than in the NTS of p53 (Figure 3). Eight hours after UV irradiation 60% of the CPD were removed from the TS of p53, and almost complete repair was observed after 24 h (Figure 3a,b). Also Msh27/7 MEFs showed pro®cient TCR of CPD in the p53 gene with kinetics comparable to those in wildtype MEFs (Figure 3) and MDFs (Figure 2). In addition, the recovery of UV-inhibited RNA synthesis following UV light (5 J/m2) was similar in both Msh27/7 MEFs and wildtype MEFs: within 3 h after UV-irradiation both cell types started to restore transcriptional activity, while almost complete recovery of RNA synthesis was reached after 9 h post-UV incubation (data not shown). Based on the in vitro experiments described above it can be concluded that the MMR gene Msh2 is not essential for TCR of UV-induced CPD in mice. The pro®cient TCR of CPD and the recovery of UVinhibited RNA synthesis is fully consistent with the

Figure 3 Repair of CPD in immortalized mouse embryonic ®broblasts. Representative autoradiogram (a) and graph (b) showing the removal of CPD from the transcribed strand (TS) and non-transcribed strand (NTS) of the active p53 gene in adenovirus E1A immortalized MEFs (Berry et al., 2000) derived from wildtype and Msh27/7 mouse embryos (de Wind et al., 1995) irradiated with 10 J/m2 UV. CPD frequencies in the 16 kb EcoRI restriction fragment of the p53 gene were measured as described in Figure 2. Msh27/7 MEFs (*, *); wildtype MEFs (&, &); TS (closed symbols); NTS (open symbols). Data points are means of at least three independent experiments and are presented as mean+s.e.m.

observation that the sensitivity of Msh27/7 mice to UV-induced erythema/edema is similar to that of wildtype mice. The wildtype level of TCR in Msh27/7 mouse cells resembles the pro®cient TCR of CPD in MMR de®cient yeast (Sweder et al., 1996), but contrasts the human MSH2 de®cient LoVo tumor cells as well as several other di€erent MMR de®cient human cell lines which show defective TCR of UVinduced CPD (Mellon et al., 1996). This discrepancy appears not to be caused by the transformed state of the human tumor cell line, since transformed Msh27/7 MEFs exhibit normal TCR of CPD like primary Msh27/7 MDFs. In addition, both types of Msh2 mutations used in this study resulted in the absence of Msh2 protein, similar to LoVo cells which are homozygous for a large deletion in the hMSH2 gene. Finally, the exponential growth rate of the LoVo tumor cells does not account for the observed TCR de®ciency, since exponentially growing wildtype and

Transcription-coupled repair in mismatch repair deficient mouse cells E Sonneveld et al

Msh27/7 MEFs show pro®cient TCR of CPD comparable with con¯uent growing MDFs. In conclusion, our data suggest that Msh2 is not essential for TCR of CPD in the mouse and that the biological e€ects of UV in Msh27/7 mice (i.e. enhanced carcinogenesis; Friedberg et al., 2000) are entirely due to the inability to repair DNA mismatches and the failure to trigger apoptosis. However, our experiments do not rule out that mouse Msh2 is essential for the TCR of oxidative damage, as shown in yeast.

Radiation Genetics and Chemical Mutagenesis, Leiden University Medical Center, Leiden, The Netherlands. Msh2D7N mice were obtained from Dr R Fodde, MGCDepartment of Human and Clinical Genetics, Leiden University Medical Center, Leiden, The Netherlands. The authors thank Dr MPG Vreeswijk for expert technical assistance and Dr N de Wind for critical reading of the manuscript. This work was supported by Dutch Cancer Society Grants UU 97-1531 and EUR 98-1800.

541

Acknowledgments The MEFs and the Msh2 antibody used in this study were obtained from Dr N de Wind, MGC-Department of References Berg RJ, Rebel H, van der Horst GT, van Kranen HJ, Mullenders LH, van Vloten WA and de Gruijl FR. (2000). Cancer Res., 60, 2858 ± 2863. Berry SE, Davis TW, Schupp JE, Hwang HS, de Wind N and Kinsella TJ. (2000). Cancer Res., 60, 5773 ± 5780. Bertrand P, Tishko€ DX, Filosi N, Dasgupta R and Kolodner RD. (1998). Proc. Natl. Acad. Sci. USA, 95, 14278 ± 14283. Bootsma D, Kraemer KH, Cleaver JE and Hoeijmakers JHJ. (1998). The Genetic Basis of Human Cancer. Vogelstein B and Kinzler KW (eds). McGraw-Hill: New York, pp. 245 ± 274. Buermeyer AB, DescheÃnes SM, Baker SM and Liskay RM. (1999). Annu. Rev. Genet., 33, 533 ± 564. Cheo DL, Ruven HJ, Meira LB, Hammer RE, Burns DK, Tappe NJ, van Zeeland AA, Mullenders LH and Friedberg EC. (1997). Mutat. Res., 4, 1 ± 9. de Laat WL, Jaspers NG and Hoeijmakers JH. (1999). Genes Dev., 13, 768 ± 785. de Wind N, Dekker M, Berns A, Radman M and te Riele H. (1995). Cell, 82, 321 ± 330. de Wind N, Dekker M, Van Rossum A, Van der Valk M and te Riele H. (1998). Cancer Res., 58, 248 ± 255. de Wind N, Dekker M, Claij N, Jansen L, van Klink Y, Radman M, Riggins G, van der Valk M, van `t Woudt K and te Riele H. (1999). Nature Genet., 23, 359 ± 362. Friedberg EC, Walker GC and Siede W. (1995). DNA repair and mutagenesis. ASM Press, Washington DC. Friedberg EC, Bond JP, Burns DK, Cheo DL, Greenblatt MS, Meira LB, Nahari D and Reis AM. (2000). Mut. Res., 459, 99 ± 108. Gowen LC, Avrutskaya AV, Latour AM, Koller BH and Leadon SA. (1998). Science, 281, 1009 ± 1012. Leadon SA and Avrutskaya AV. (1997). Cancer Res., 57, 3784 ± 3791. Leadon SA and Avrutskaya AV. (1998). Mut. Res., 407, 177 ± 187.

Liu B, Nicolaides NC, Markowitz S, Willson JKV, Parsons RE, Jen J, Papadopoulos N, PeltomaÈki P, de la Chapelle A, Hamilton SR, Kinzler KW and Vogelstein B. (1995). Nature Genet., 9, 48 ± 55. Lynch HT, Smyrk T and Lynch J. (1997). Cancer Genet. Cytogenet., 93, 84 ± 99. Mellon I, Rajpal DK, Koi M, Boland C and Champe GN. (1996). Science, 272, 557 ± 560. Mellon I and Champe GN. (1996). Proc. Natl. Acad. Sci. USA, 93, 1292 ± 1297. Reitmar AH, Risley R, Bristow RG, Wilson T, Ganesh A, Jang A, Peacock J, Benchimol S, Hill RP, Mak TW, Fishel R and Meuth M. (1997). Cancer Res., 57, 3765 ± 3771. Rey O, Lee S and Park NH. (1999). Oncogene, 18, 6997 ± 7001. Smits R, Ho¯and N, Edelmann W, Geugien M, JagmohanChangur S, Albuquerque C, Breukel C, Kucherlapati R, Kielman MF and Fodde R. (2000). Genes Chromosom. Cancer, 29, 229 ± 239. Sweder KS, Verhage RA, Crowley DJ, Crouse GF, Brouwer J and Hanawalt PC. (1996). Genetics, 143, 1127 ± 1135. van Ho€en A, Natarajan AT, Mayne LV, van Zeeland AA, Mullenders LH and Venema J. (1993). Nucl. Acids Res., 21, 5890 ± 5895. van Ho€en A, Venema J, Meschini R, van Zeeland AA and Mullenders LHF. (1995). EMBO J., 14, 360 ± 367. van Oosten M, Rebel H, Friedberg EC, van Steeg H, van Der Horst GT, van Kranen HJ, Westerman A, van Zeeland AA, Mullenders LH and de Gruijl FR. (2000). Proc. Natl. Acad. Sci. USA, 97, 11268 ± 11273. van Sloun PPH, Jansen JG, Weeda G, Mullenders LHF, van Zeeland AA, Lohman PHM and Vrieling H. (1999). Nucl. Acids Res., 27, 3276 ± 3282. Wu J, Gu L, Wang H, Geacintov NE and Li G-M. (1999). Mol. Cell. Biol., 19, 8292 ± 8301.

Oncogene